Introduction:
Large cities like Dallas have historically struggled with crime, so the Center for Policing Equity (CPE) and police agencies across teamed up to produce open-source crime data. This enables data analysts to find trends and produce insightful information for decision-makers to enhance public safety. Working in this field is quite dangerous due to its nature. By examining data on injuries sustained by police officers in Dallas, we can obtain useful insights about the risks involved and identify potential improvements.
In this project and with the help of open-source data from Kaggle containing information about the different types of incidents in Dallas during the year 2016-2017, we will determine which instances are more likely to result in an injury or hospitalization. This information might help identify dangerous encounters and prepare in advance to handle them.
The data used in this project is available here: https://www.kaggle.com/datasets/center-for-policing-equity/data-science-for-good.
The police force in Dallas is comprised of 2140 officers, of whom 89.93% are male and only 10.07% are female, showing a great gender imbalance ratio.
OFFICER_GENDER
Female Male
10.07 89.93
When we look at the proportion of injuries by gender, 12.92% of female officers were injured compared to 9.47% of their male counterparts. This might suggest that female cops are at greater risk of getting injured while performing their duties.
Next, we explore if there is a connection between injuries and officer ethnicity. According to the graph, white officers represent the majority in the police force in Dallas and also represent the most cases of injuries.
However, when comparing proportions, it seems that Asian officers and officers of other ethnicities, with 14.81% and 14.55% respectively, are more likely to get injured, while Native American officers are the least likely to do so.
OFFICER_INJURY
OFFICER_RACE No Yes
American Ind 100.00 0.00
Asian 85.45 14.55
Black 90.91 9.09
Hispanic 90.87 9.13
Other 85.19 14.81
White 90.00 10.00
Now that we have a clearer picture of injuries by gender and ethnicity among police staff, what are the main injuries that police officers sustain? The analysis shows that 90.64% of injuries are not noted or visible, while the rest fall into the category of minor injuries. These results suggest the data might be incorrect or that officers have pre-existing health issues such as diabetes or heart problems.
OFFICER_INJURY_TYPE
No injuries noted or visible Abrasion/Scrape
90.64 3.06
Laceration/Cut Sprain/Strain
0.59 0.59
Redness/Swelling Fluid Exposure
0.55 0.42
Perhaps, there is an effect in years on officer experience with number of injuries or hospitalizations. In the graph we can observe a big amount of officers staying in the department for three years. Examining the work experience in the police department, it was found that the median is 6 years and that a great amount of police officers stay in the department for the first three years. However, in the Five year mark there is a significant dip in the number of police officers, perhaps this is the year when the majority of police officers choose to change careers or remain in the field.
Also on average, Native American officers stay in the force for around 16 years, Black officers for around 7 years, White, Hispanic, and other officers for about 6 years, and Asian officers stay for the shortest period of time, about 3 years.
Summary of years on force Min. 1st Qu. Median Mean 3rd Qu. Max.
0.000 3.000 6.000 8.049 10.000 36.000
Summary of years on force for injured officers
Min. 1st Qu. Median Mean 3rd Qu. Max.
1.000 3.000 7.500 9.526 11.750 36.000
Summary of years on force for hospitalized officers
Min. 1st Qu. Median Mean 3rd Qu. Max.
1.000 3.000 6.500 9.562 14.000 31.000
In addition, injured officers stay in the force for a median of 7.5 years, while hospitalised officers stay for 6.5 years. This supports the idea that the longer the time spent in the force, the higher the chance to get injured or hospitalised.
The main reasons for using force are: arrest, danger to oneself or others, and active aggression. To explore the possible connection between the use of force and officer injury a correlation analysis was developed.
The analysis takes into account the officers’ gender, the officer’s injury, the officer’s hospitalisation, the subject’s injury, and the subject’s arrest. Surprisingly, no strong correlation was found between these variables. The only moderate correlation was found between officer injury and officer hospitalisation, meaning that suffering an injury increases the chances of ending up hospitalised.
Risky situations such as active aggression, aggressive animals, and danger to others and to oneself might be associated with an officer’s injuries. In the scatter plot, we can observe the exact location of each of these incidents.
In general, we can see that most points concentrate in the middle of the plot, meaning that most dangerous calls tend to occur in downtown Dallas.
Since risky situations such as active aggression, aggressive animals, and danger to others and to oneself might be associated with an officer’s injuries, an interactive map of Dallas showing all instances that resulted in an injured or hospitalized officer was created. Green points represent instances where officers got injured, while red points represent instances where officers got hospitalized.
It was found that the majority of the points were concentrated near the center of the chart, representing downtown.
Arrest Active Aggression Assault to Other Person
22 18 4
Danger to self or others Detention/Frisk Aggressive Animal
3 1 0
Barricaded Person Crowd Disbursement NULL
0 0 0
Other Property Destruction Weapon Display
0 0 0
Most cases of hospitalized officers were due to arrests, followed by Active Aggression, Assault to Other Person, Danger to self and others, and Detention/Frisk. Surprisingly, no cases of hospitalization were due to aggressive animals or weapon display.
Using a time series plot revealed that the majority of incidents occurred between February-March and October, while November had the fewest number of incidents.
If we plot a trend line, we can observe that the number of crimes showed a declining trend during the year 2016-2017.
In conclusion, after analysing the possible causes of injuries and hospitalisations of police officers in Dallas, Texas, it was found that there is a great gender imbalance in the force and that female officers are at greater risk of getting injured than their male counterparts.
Also, when comparing ethnicities, it was found that the majority of officers were white, showing room for improvement in diversity and representation. In addition, Other and Asian officers are the ones more likely to get injured, even though they are part of the minority in the police department. Most injuries among officers were not visible, suggesting a problem with reporting methods or pre-existing health problems.
In the analysis, it is inferred that using force resulted in more injuries and hospitalisations, especially around downtown Dallas when officers arrested subjects and attended calls of active aggression, due to the dangerous nature of handling these types of crimes. The decline in crime during 2016–2017 could suggest that crime prevention policies in Dallas, Texas, were successful. However, to find the true causes of this declining trend, further research is necessary.
This exploratory analysis has its limitations and shortfalls due to its nature; to correctly draw conclusions, further research is necessary.
Bibliography:
Holtz, Yan. “R Color Brewer’s Palettes.” – The R Graph Gallery, https://r-graph-gallery.com/38-rcolorbrewers-palettes.html?utm_content=cmp-true.
Alboukadel. “Top R Color Palettes to Know for Great Data Visualization.” Datanovia, 25 Dec. 2019, https://www.datanovia.com/en/blog/top-r-color-palettes-to-know-for-great-data-visualization/.
“Levels: Levels Attributes.” RDocumentation, https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/levels.